Neoadjuvant Immunotherapy and Chemoimmunotherapy Regimens in Head and Neck Cancer: A Systematic Review and Meta-Analysis

Here is an expanded summary of Baratz et al., “Neoadjuvant Immunotherapy and Chemoimmunotherapy Regimens in Head and Neck Cancer: A Systematic Review and Meta-Analysis,” published online in JAMA Otolaryngology–Head & Neck Surgery on March 12, 2026

• Why this paper matters:
  • Neoadjuvant immunotherapy in resectable HNSCC:
    • Has been attractive because:
      • It may treat micrometastatic disease early
      • Exploit the intact tumor-immune microenvironment before surgery
      • It may potentially improve pathologic response without delaying definitive treatment
    • What has been unclear is whether immunotherapy alone or chemoimmunotherapy is more effective
  • This meta-analysis addresses that question by pooling the available prospective data in resectable, treatment-naive HNSCC
• Objective:
  • The investigators aimed to:
    • Summarize the efficacy of neoadjuvant chemoimmunotherapy in HNSCC
    • Compare outcomes of chemoimmunotherapy vs immunotherapy alone:
      • Before definitive surgery in locoregionally advanced resectable HNSCC
• Methods:
  • This was a librarian-led systematic review and meta-analysis performed according to PRISMA methodology
  • The authors searched MEDLINE, EMBASE, Cochrane Central, Cochrane Database of Systematic Reviews, and Scopus from database inception through October 2024
  • They included prospective interventional trials in resectable, treatment-naive HNSCC that had completed accrual and reported pathologic response and / or RECIST response data
  • Two investigators independently performed study screening and extraction
  • The main outcomes were:
    • Major pathologic response (MPR)
    • Complete pathologic response (CPR)
    • Complete radiographic response (CR):
      • By RECIST 1.1
  • Secondary outcomes included:
    • 1-year overall survival
    • Toxicity
  • The pooled analysis used a binary random-effects model, with heterogeneity reported as I² 
  • Study population:
    • The meta-analysis included 23 studies with a total of 751 patients. Of these:
      • 357 patients (47%) received chemoimmunotherapy
      • 102 patients (14%) received dual-agent immunotherapy
      • 292 patients (39%) received single-agent immunotherapy 
    • The pooled cohort was predominantly male (77%) with an age range of 27 to 87 years 
• Main findings:
  • Pathologic response:
    • Strongly favored chemoimmunotherapy
  • The most important finding was:
    • The marked gradient in pathologic response across regimens:
      • Pooled MPR + CPR rates were:
        • 66% for chemoimmunotherapy 95% CI 58%-73%
        • 18% for dual-agent immunotherapy 95% CI 6%–29%
        • 6% for single-agent immunotherapy 95% CI 3%–9% 
  • This is the key take-home point:
    • Adding chemotherapy to immunotherapy:
      • Was associated with substantially higher pathologic response rates than immunotherapy alone
    • Clinically, this matters because in head and neck cancer:
      • Pathologic response has increasingly been explored as an early signal of antitumor activity and a possible surrogate for longer-term benefit:
        • Although it is not yet a fully validated surrogate for survival in this setting
        • That distinction is important when interpreting these results
    • The paper shows better tumor kill in the surgical specimen:
      • But it does not yet prove that patients live longer because of the neoadjuvant regimen:
      • That is why the authors call for phase 3 trials
  • Short-term survival looked promising across groups, but differences were not definitive:
    • Across the included studies, 1-year overall survival ranged:
      • 88% to 96% with single-agent immunotherapy
      • 88% to 96% with dual-agent immunotherapy
      • 88% to 100% with chemoimmunotherapy 
    • These ranges suggest that all three strategies can be delivered with good short-term outcomes in selected patients:
      • However, because these were mainly early-phase, non-comparative studies with heterogeneous populations and follow-up:
        • The survival data should be viewed as hypothesis-generating, not practice-defining
  • Toxicity was higher than dual immunotherapy, but not prohibitive:
    • Among studies reporting adverse events, grade 3 to 5 adverse events occurred in:
      • 29% of patients receiving single-agent immunotherapy
      • 3% with dual-agent immunotherapy
      • 17% with chemoimmunotherapy 
    • These numbers need cautious interpretation because toxicity reporting was not uniform across studies, and the denominators were limited to reporting studies rather than all pooled patients:
      • Still, the overall message is that chemoimmunotherapy increased efficacy while maintaining an acceptable:
        • Though not trivial, toxicity burden in selected surgical candidates. 
  • Authors’ conclusion:
    • The authors concluded that neoadjuvant chemoimmunotherapy:
      • Was associated with higher pathologic and radiographic response rates than immunotherapy alone in locoregionally advanced resectable HNSCC, and that these findings support the need for head-to-head phase 3 trials
• How to interpret this as a head and neck oncologic surgeon:
  • Strengths:
    • This study has several strengths:
      • It focuses specifically on resectable, treatment-naive HNSCC:
        • Which is the clinically relevant population for neoadjuvant decision-making
      • It includes only prospective interventional studies
      • It separates single-agent, dual-agent, and chemoimmunotherapy approaches rather than lumping all neoadjuvant immunotherapy together
      • It uses outcomes surgeons and multidisciplinary teams care about:
        • Pathologic response, radiographic response, survival, and toxicity
  • Important limitations:
    • The paper is very useful, but it does not settle the question of standard of care
    • The biggest limitations are:
      • Most included studies were phase 1 / 2, small, and often single-arm
      • There was likely substantial clinical heterogeneity:
        • Primary site, stage, PD-L1 status, regimen, number of cycles, and adjuvant treatment strategies
      • The outcome driving the signal is primarily pathologic response:
        • Not mature event-free survival or overall survival
      • Cross-trial comparisons may exaggerate differences:
        • Because these were not randomized head-to-head comparisons
      • Toxicity and imaging response reporting were not fully standardized
• So the paper supports promise:
  • Not final proof
• Practical clinical implications:
  • For a practicing surgeon:
    • This meta-analysis suggests that chemoimmunotherapy is currently the most active neoadjuvant immune-based strategy in resectable HNSCC:
      • At least if the endpoint is pathologic response
  • If a center is considering neoadjuvant treatment within a trial or highly selected multidisciplinary framework:
    • The data support prioritizing chemoimmunotherapy over immunotherapy alone when the goal is maximizing preoperative tumor regression
  • At the same time, these data do not mean every resectable oral cavity, larynx, or oropharynx patient:
    • Should routinely receive neoadjuvant chemoimmunotherapy outside a protocol
  • The field is moving quickly, and the editorial accompanying this paper emphasizes that these results arrive in the context of KEYNOTE-689:
    • The first phase 3 randomized study to establish perioperative immunotherapy as a standard-of-care option in locally advanced resectable HNSCC:
      • While also warning that enthusiasm should be balanced with caution as these strategies enter broader practice
• Bottom line:
  • This meta-analysis is one of the clearest pooled signals so far that in resectable locoregionally advanced HNSCC, neoadjuvant chemoimmunotherapy produces substantially higher pathologic response rates than immunotherapy alone
  • The benefit signal is strong for tumor response, short-term survival appears encouraging, and toxicity seems manageable in selected patients:
    • But the evidence base is still dominated by early-phase studies, so phase 3 randomized data remain essential before universal adoption